Is this a bug in decode function?????

Hi,
I'm trying run the following query and it blows with a ORA-01722. I'm dealing with the sample HR schema that comes with the database. The SQL is :
select employee_id, start_date, end_date, department_id,
decode(job_id, 'IT_PROG', 1,'AC_ACCOUNT', 2,'AC_MGR', 3,'MK_REP', 4,'ST_CLERK', 5,'AD_ASST', 6,'SA_REP', 7,'AC_ACCOUNT', 8,'other') Job_num, job_id
from job_history
blows up. This is what I get:
ORA-01722: invalid number
But strangely the below query works:
select employee_id, start_date, end_date, department_id,
decode(job_id, 'IT_PROG', '1','AC_ACCOUNT', '2','AC_MGR', '3','MK_REP', '4','ST_CLERK', '5','AD_ASST', '6','SA_REP', '7','AC_ACCOUNT', '8', 9999)Job_num, job_id
from job_history
The result of the second query in a comma delimited format is as below:
EMPLOYEE_ID,START_DATE,END_DATE,DEPARTMENT_ID,JOB_NUM,JOB_ID
102,1/13/1993,7/24/1998,60,1,IT_PROG
101,9/21/1989,10/27/1993,110,2,AC_ACCOUNT
101,10/28/1993,3/15/1997,110,3,AC_MGR
201,2/17/1996,12/19/1999,20,4,MK_REP
114,3/24/1998,12/31/1999,50,5,ST_CLERK
122,1/1/1999,12/31/1999,50,5,ST_CLERK
200,9/17/1987,6/17/1993,90,6,AD_ASST
176,3/24/1998,12/31/1998,80,7,SA_REP
176,1/1/1999,12/31/1999,80,9999,SA_MAN
200,7/1/1994,12/31/1998,90,2,AC_ACCOUNT
Any idea why this happens? I'm going bonkers. Your help is greatly appreciated.
Best Regards,
Naveen.

It's a feature :) Databases generally provide type coercion for common data types assuming that the user actually meant the same value but as a different type. Hence, '1' -> 1. However, you couldn't directly convert 'A' to a relevant number as it could mean 10 in hex, 65 as an ASCII value, etc.

Similar Messages

Help on Decode function in Toplink

Hi,
I have a sql statement like this:
select field1, field2 from table1 where upper(field1) like 'ADELPHI%'
order by field1,
DECODE(TO_CHAR(NVL(LENGTH(TRANSLATE(field1,'A1234567890','A')),0)),'0',LPAD(field1 ,50),field1)
How would I converted this to toplink query?
I appreciate any help you guys can extend to me.

I just spent about a half hour writing a response here, and the damned forum timed out on me. #$^!@$#!^!@$%.
Here is the summary:
- Using TL Expressions and EJBQL are to allow Java developers who don't necessarily know SQL to write queries and to facilitate portability/maintainability of the apps. For REALLY gnarly queries, using TL Expressions or EJBQL won't buy you this. I.e., SQL is more expressive and easier for this kind of quer, so it's not easier to use, and frankly with the code needed to support it, it won't be easier to maintain or port... Bottom line, maybe SQL is the best option here?
- If you want to experiment, search for TO_CHAR on this forum, and the decode function in the javadoc. You need to build the expressions in pieces and you can treat expressions as arguments to other expressions -- in other words, you can build an expression like builder.leftPad(50, builder.get("field1")) and use it as one of the key/values in the hashtable to the decode function (in theory).
- Don

How can i use this expression in DECODE function?

My PNO table formart is
PNO PDate PCount
P001 08/27/05 09:45 20
P001 08/27/05 09:50 10
P002 08/27/05 03:40 20
P003 08/28/05 11:00 20
P003 09/28/05 10:00 20
P003 08/27/05 11:00 10
P003 09/27/05 04:00 50
I want to display total pcount for shift1(morning) and shift2(evening) & the date is 08/27/05
PNO Shift1(08/27/2005 08:00 to 12:30) Shift2(08/27/2005 02:00 to 05:30)
P001 30 0
P002 0 20
P004 10 50
So for shift1 condition is:
pdate>=08/27/2005 08:00 and pdate<=08/27/2005 12:30
So for shift2 condition is:
pdate>=08/27/2005 02:00 and pdate<=08/27/2005 05:30
If i tried to give this expression in decode function , its not working.
Please give me the sql query to solve this problem.
Millions of thanks in advacne

forgot to paste the output for a given date.
SQL> create table pno
2 (pno    varchar2(32)
3 ,pdate date
4 ,pcount number)
5 /
Table created.
SQL>
SQL> insert into pno values('P001', to_Date('08/27/05 09:45', 'mm/dd/yy hh:mi') ,20)
2 /
1 row created.
SQL> insert into pno values('P001', to_Date('08/27/05 09:50', 'mm/dd/yy hh:mi') ,10)
2 /
1 row created.
SQL> insert into pno values('P002', to_Date('08/27/05 03:40', 'mm/dd/yy hh:mi') ,20)
2 /
1 row created.
SQL>
SQL> insert into pno values('P003', to_Date('08/28/05 11:00', 'mm/dd/yy hh:mi') ,20)
2 /
1 row created.
SQL> insert into pno values('P003', to_Date('09/28/05 10:00', 'mm/dd/yy hh:mi') ,20)
2 /
1 row created.
SQL> insert into pno values('P003', to_Date('08/27/05 11:00', 'mm/dd/yy hh:mi') ,10)
2 /
1 row created.
SQL> insert into pno values('P003', to_Date('09/27/05 04:00', 'mm/dd/yy hh:mi') ,50)
2 /
1 row created.
SQL> select pno,sum(pcount) "Morning Shift Count"
2 from pno
3 where to_char(pdate, 'mm/dd/yyyy hh:mi') >= '08/27/2005 08:00'
4    and to_char(pdate, 'mm/dd/yyyy hh:mi') <='08/27/2005 12:30'
5 group by pno
6 /
PNO                              Morning Shift Count
P001                                              30
P003                                              10
SQL> select pno,sum(pcount) "Evening Shift Count"
2 from pno
3 where to_char(pdate, 'mm/dd/yyyy hh:mi') >= '08/27/2005 02:00'
4    and to_char(pdate, 'mm/dd/yyyy hh:mi') <='08/27/2005 05:30'
5 group by pno
6 /
PNO                              Evening Shift Count
P002                                              20

Decode Function I asked this in the wrong section so I hope Im right now!

select
a11.SEQ_MEMB_ID,
a11.Auth_Number,
a11.Admit_Primary_Date,
a11.Reviewer,
Decode (a11.Reviewer,
CLC, '2',
HBR, '3',
RAB, '4')
a11.admit_primary_date
from windsoradm.auth_master a11;
I am trying to get if the Reviewer is CLC then they are code 2 HBR Code 3 etc and it does not work. I take out the decode function and the query works. I am getting ORA-00923 FROM KEYWORD NOT FOUND WHERE EXPECTED.

Hi,
Whenever you have a problem, please post a little sample data (CREATE TABLE and INSERT statements, relevant columns only) from all tables involved.
Also post the results you want from that data, and an explanation of how you get those results from that data, with specific examples.
Simplify the problem as much as possible. Remove all tables and columns that play no role in this problem.
Always say which version of Oracle you're using.
This is the right forum for this question. Mark your question in the other forum {message:id=10266930} as "Answered", so people won't waste their time.
925518 wrote:
select
a11.SEQ_MEMB_ID,
a11.Auth_Number,
a11.Admit_Primary_Date,
a11.Reviewer,
Decode (a11.Reviewer,
CLC, '2',
HBR, '3',
RAB, '4')
a11.admit_primary_date
from windsoradm.auth_master a11;
I am trying to get if the Reviewer is CLC then they are code 2 HBR Code 3 etc and it does not work. I take out the decode function and the query works. I am getting ORA-00923 FROM KEYWORD NOT FOUND WHERE EXPECTED.That's the correct DECODE syntax, where CLC, HBR and RAB are columns, and you want to return a string (not a number) such as '2'.
If 'CLC', 'HBR' and 'RAB' are literal values, then enclose them in single-quotes. If you want to return a number, don't put it in quotes. (Anyhting inside single-quotes is a string literal.)
Perhaps you meant something like:
,     DECODE ( a11.Reviewer
            , 'CLC'          , 1
            , 'HBR'          , 2
            , 'RAB'          , 3
            )      AS reviewer_codeThere's an error right after the DECODE. You can't use a11.admit_primary_date as a column alias. (This is probably what's causing the ORA-00923 error.) Either a11_admit_primary_date (with an underscore instead of a dot) or admit_primary_date would be okay. If a11.admit_priomary_date is the next column in the SELECT clause, put a comma befor it.
You need a FROM clause.

Wat is wrong in this decode function, it doesn get executed,please help me

wat is wrong in this decode function, it doesn get executed,please help me
decode(DVI.COMPANY||'.')||
                                        DECODE(DVI.COST_CENTRE||'.')||
                                        DECODE(DVI.ACCOUNT||'.')||
                                        DECODE(DVI.LOCATION||'.')||
                                        DECODE(DVI.PRODUCT||'.')||
                                        DECODE(DVI.FUTURE)company_id,

Hi,
I could not understand what you are trying to do with decode here.
The standard way of using the decode is like :
SQL>SELECT DECODE (1 , 1 ,'TRUE','FALSE') FROM DUAL;
DECO
TRUE
1 row selected.
1* SELECT DECODE (1 , 0 ,'TRUE','FALSE') FROM DUAL
SQL>/
DECOD
FALSE
1 row selected.
Please explain what you want to do with DECODE ?
Regards

DECODE Function In Reports 3.0

Hi,
When I use the decode function in a query (Reports 3.0.5.8.0) I receive an error message "Bind Variable Does Not Exist" (ORA-1006). When I remove the decode function from the query, the query compiles fine. All table and column references are existing.
Also, the same query runs perfectly fine from SQL*Plus. The OS is Win 2000, SP2
Any idea if this is a bug? Is there any way out? Do get back as this is very urgent.
The database is Oracle 8i Ent, 8.1.7 on AIX 4.3
The query is as follows:
SELECT NAM_CMP , COD_CMP , COD_CMP_REG ,
DECODE(FLG_CMP_SHP,'Y','S/',' /') || DECODE(FLG_CMP_AGT,'Y','A/',' /') || DECODE(FLG_CMP_CNE,'Y','C',' ') TYPE ,
DES_ADD_STT , DES_ADD_PLC , DES_ADD_STA , DES_ADD_POS , COD_CTY , COD_cou_iso ,
DAT_LST_AST , DAT_LST_UPD,
cod_dcl_box_num, decode(cod_imp_chg_way,'C','CASH','CREDIT') cod_imp_chg_way,
decode(cod_imp_cre_way,'W','WEEKLY','F','FIFTEEN DAYS','M','MONTHLY', NULL) cod_imp_cre_way,
decode(flg_imp_prn_inv, 'N', null, flg_imp_prn_inv) flg_imp_prn_inv,
cod_bro_box_num, decode(cod_exp_chg_way,'C','CASH','CREDIT') cod_exp_chg_way,
decode(cod_exp_cre_way,'W','WEEKLY','F','FIFTEEN DAYS','M','MONTHLY', NULL) cod_exp_cre_way,
decode(flg_exp_prn_inv, 'N', null, flg_exp_prn_inv) flg_exp_prn_inv
FROM CMP
where dat_can is null

Try TO_CHAR(NULL) instead of NULL in all DECODE functions.
For example:
decode(cod_imp_cre_way,'W','WEEKLY','F','FIFTEEN DAYS','M','MONTHLY', TO_CHAR(NULL))
instead of:
decode(cod_imp_cre_way,'W','WEEKLY','F','FIFTEEN DAYS','M','MONTHLY', NULL).

How to use DECODE function in Exspression?

Hi,
Can we use DECODE in Expression?
I'm trying to use DECODE function but there is an error during the validation. But when i validate the mapping, it is successfully compiled but it is failed during deployment.
But if I use CASE instead of DECODE, it works fine.
Can we use DECODE in OWB???
Thanks
Raj

Hi,
In OWB 10gR2, if your are using only one DECODE in an expression, it's working. The package will compile when deploying the mapping. OWB will replace the DECODE by a CASE.
But when you are using nested decode in an expression ( for example : decode(col1, 1, 'M', decode(col2, 'Madame', 'Mme', null)) ) only the first one is replaced by a case at deployment.
In ROW_BASED mode, text of the expression is used outside of an sql statement and deployment will fails with "PLS-00204: function or pseudo-column 'DECODE' may be used inside a SQL statement only."
If operating mode for the mapping is set to SET_BASED, it's working because the expression is used in an sql statement.
I have logged a SR in metalink for this issue and a bug is opened (bug 5414112).
But I agree with you, it's better to use case statement.
Bernard

Error -3016 when trying to compare a column value via "decode" function

The following statement:
select * from turbine_user
where login_name = decode(1, 2, 'name1', 'name2')
yields a "-3016 Invalid numeric constants" error.
OTOH, the following variations work as expected:
(a)
select * from turbine_user
where login_name = 'name1'
(b)
select * from turbine_user
where login_name = 'name2'
(c)
select * from turbine_user
where 'name1' = decode(1, 2, 'name1', 'name2')
(d)
select decode(1, 2, 'name1', 'name2') from turbine_user
Does MaxDB have problems with using the decode function to compare something to column values?
My MaxDB-Version: 7.6.03.07
Thanks in advance,
Alex
Edited by: Alexander Zimmer on Feb 6, 2008 3:53 PM

Hi,
unfortunately I assume that I should explain the doing of decode a little bit:
As we see in the reference manual:
DECODE(<check_expression>,<search_and_result_spec>,...[,<default_expression>])
<check_expression> ::=
<expression>
<search_and_result_spec> ::=
<search_expression>,<result_expression>
<default_expression> ::=
<expression>
<search_expression> ::=
<expression>
<result_expression> ::=
<expression>
Ok, that will mean, that the first parameter of decode will be compared to parameter number 2, 4, 6, 8, 10,... up to number n-1.
If the result of the comparison is true with parameter number x, the result of decode will be the value of parameter x+1.
If no comparison results in 'equal', then NULL will be returned if there is an odd number of parameters respectively the value of the last parameter if there is an even number of parameters of the decode function.
In your context, this means, that 1 is compared to 2 (always false), thus returning 'name2' always. Therefore I do not understand what you want to do with that decode.
Ok, if we assume that the select sent is only a short form of what you really want to do, then I have to apologize, but there is a bug in the kernel coding. It will be fixed ASAP.
To be able to work until then, put the function DIGITS around your 1 and around the 2 (or the corresponding real parameter values) and it should do.
BTW: with column on the left hand side, the error occurs and with a string literal it does not? This is different to my check where both values on the left side cause this error to occur.
Elke

SQL DECODE function

Oracle version: 9.2.0.1.0
Procob version: 9.2.0.1.0
Cobol version: 3.1.11
O.S. version: Windows XP
We have a pro*cobol file with the next conversion rule into a select defined with a cursor.
NVL(TO_CHAR(A.F_ALTA,'YYYYMMDD'),' '),
NVL(TO_CHAR(A.F_ESTADO,'YYYYMMDD'),' '),
we store the result data through a fetch of the cursor into two variables (W-F-ALTA and W-F-ESTADO) When the program executes an insert of these two values in a target table, we define the next conversion rule:
TO_DATE(DECODE(:W-F-ALTA,' ',NULL,:W-F-ALTA),'YYYYMMDD'),
TO_DATE(DECODE(:W-F-ESTADO,' ',NULL,:W-F-ESTADO),'YYYYMMDD')
Pro*Cobol and Cobol does not have errors when we precompile and compile the program but, when we execute the generated exe file, Oracle displays the next error:
ORA-01841 (full) year must be between -4713 and +9999, and not be 0
We think that the DECODE function is not understood by the cobol or pro*cobol interpreter. If we execute the same code without the DECODE function (only with TO_DATE function) there is no errors in the execution of the program
somebody knows if there are reported problems with this DECODE function in Pro*Cobol or Cobol and how to solve them?
Thanks

TO_DATE(DECODE(:W-F-ALTA,' ',NULL,:W-F-ALTA),'YYYYMMDD'),
TO_DATE(DECODE(:W-F-ESTADO,' ',NULL,:W-F-ESTADO),'YYYYMMDD')The problem is with the input values of :W-F-ALTA and :W-F-ESTADO. Using the first line as an example, what you are saying is if :W-F-ALTA is ' ', substitute it with NULL, any other value should be :W-F-ALTA, then output the TO_DATE value of whatever the result is.
So if the value of :W-F-ALTA is the wrong charater or in the wrong input format, TO_DATE function does not understand it.
See examples of success and failures below with your Bind variables substituted with typical values.
SQL> select TO_DATE(DECODE(' ',' ',NULL,'20060830'),'YYYYMMDD') from dual;
TO_DATE(D
SQL> select TO_DATE(DECODE('1999-01-01',' ','NULL','20060830'),'YYYYMMDD') from dual;
TO_DATE(D
30-AUG-06
SQL> select TO_DATE(DECODE('SHOULD_ERROR',' ','NULL','SHOULD_ERROR'),'YYYYMMDD') from dual;
select TO_DATE(DECODE('SHOULD_ERROR',' ','NULL','SHOULD_ERROR'),'YYYYMMDD') from dual
ERROR at line 1:
ORA-01841: (full) year must be between -4713 and +9999, and not be 0
SQL> select TO_DATE(DECODE('20060830',' ','NULL','20060830'),'YYYYMMDD') from dual;
TO_DATE(D
30-AUG-06
SQL> select TO_DATE(DECODE('12345678',' ','NULL','12345678'),'YYYYMMDD') from dual;
select TO_DATE(DECODE('12345678',' ','NULL','12345678'),'YYYYMMDD') from dual
ERROR at line 1:
ORA-01843: not a valid month
SQL>

Problem using DECODE() function with a Query of Queries

I
posted
on my blog about an issue I was having trying to use the PL/SQL
DECODE() function with a Coldfusion Query of Queries. This function
works fine when you query a database for information. However, when
you query another query, it seems that CF doesn't recognize it. I
got errors stating that it found a left parenthesis where it
expected a FROM key word. Here is a simplified version of what I am
trying to do:
quote:

<cfscript>
qOriginal = queryNew("Name,Email,CountryCode",
"VarChar,VarChar,VarChar");
newRow = queryAddRow(qOriginal, 5);
querySetCell(qOriginal, "Name", "Joe", 1);
querySetCell(qOriginal, "Email", "[email protected]", 1);
querySetCell(qOriginal, "CountryCode", "AMER", 1);
querySetCell(qOriginal, "Name", "Sally", 2);
querySetCell(qOriginal, "Email", "[email protected]", 2);
querySetCell(qOriginal, "CountryCode", "AMER", 2);
querySetCell(qOriginal, "Name", "Bob", 3);
querySetCell(qOriginal, "Email", "[email protected]", 3);
querySetCell(qOriginal, "CountryCode", "ASIA", 3);
querySetCell(qOriginal, "Name", "Mary", 4);
querySetCell(qOriginal, "Email", "[email protected]", 4);
querySetCell(qOriginal, "CountryCode", "EURO", 4);
querySetCell(qOriginal, "Name", "John", 5);
querySetCell(qOriginal, "Email", "[email protected]", 5);
querySetCell(qOriginal, "CountryCode", "EURO", 5);
</cfscript>
<cfquery name="qCountries" dbtype="query">
SELECT DISTINCT(CountryCode) AS CountryCode,
DECODE(states, "AMER", "North America & Canada",
"EURO", "Europe & Africa", "ASIA", "Japan &
Asia","") CountryName
FROM qOriginal
ORDER BY CountryCode
</cfquery>
<cfdump var="#qCountries#">

So running this returned the following error:
Query Of Queries syntax error.
Encountered "(. Incorrect Select Statement, Expecting a
'FROM', but encountered '(' instead, A select statement should have
a 'FROM' construct.
Does anybody know why this doesn't work? Is it just not
supported? Please note that I have also tried to use the CASE()
function instead of DECODE() and that resulted in basically the
same error. For now I an looping over my distinct query with a
switch statement and manually loading a new query with the data how
I want it. But it would be a lot cleaner and less code to have the
DECODE() to work. Thx!

DECODE() is an Oracle function, not generic SQL. Q-of-Q is a
very limited subset of SQL and lacks many functions and clauses
available in standard SQL, especially what you may be used to using
in your particular RDBMS.
See
Query
of Queries user guide
Phil

Regarding decode function

Hi all,
i want to know abt decode function in oracle-sql
i..e passing the parameters in decode using decode using ":" bind parameter
select JOB, decode(:j,'CLERK','MANAGER','ANALYST','EXEC',JOB) FROM EMP;
i declared the j variable in sql environment
VARIABLE J VARCHAR2(20);
i exec the query
and passed as the clerk as input parameter
but iam getting the o/p as
JOB DECODE(:J
MANAGER MANAGER
MANAGER MANAGER
MANAGER MANAGER
SALESMAN SALESMAN
SALESMAN SALESMAN
SALESMAN SALESMAN
CLERK CLERK
SALESMAN SALESMAN
ANALYST ANALYST
CLERK CLERK
ANALYST ANALYST
JOB DECODE(:J
CLERK CLERK
12 rows selected.
SQL> PRINT J;
J
im not getting o/p
so please hlelp me only : should be used not &
this is my first thread.
execuse me if any mistakes
tons of thanks in advance to all

SORRY DAVE,ROD FOR NOT BEING CLEAR,
IAM WORKING WITH EMP TABLE THE DATA IN IT IS LIKE THIS
EMPNO ENAME JOB MGR HIREDATE SAL COMM DEPTNO
7698 BLAKE MANAGER 7839 01-MAY-81 2850 30
7782 CLARK MANAGER 7839 09-JUN-81 2450 10
7566 JONES MANAGER 7839 02-APR-81 2975 20
7654 MARTIN SALESMAN 7698 28-SEP-81 1250 1400 30
7499 chaitu SALESMAN 7698 20-FEB-81 1600 300 10
7844 TURNER SALESMAN 7698 08-SEP-81 1500 0 30
7900 JAMES CLERK 7698 03-DEC-81 950 30
7521 WARD SALESMAN 7698 22-FEB-81 1250 500 30
7902 FORD ANALYST 7566 03-DEC-81 3000 20
7369 SMITH CLERK 7902 17-DEC-80 800 20
7788 SCOTT ANALYST 7566 09-DEC-82 3000 20
EMPNO ENAME JOB MGR HIREDATE SAL COMM DEPTNO
7876 ADAMS CLERK 7788 12-JAN-83 1100 20
AND I USED DECODE FUNCTION
select JOB BEFOREPROMOTION, decode(job,'CLERK','MANAGER','ANALYST','EXEC') PROMOTION FROM EMP
BEFOREPRO PROMOTI
MANAGER
MANAGER
MANAGER
SALESMAN
SALESMAN
SALESMAN
CLERK MANAGER
SALESMAN
ANALYST EXEC
CLERK MANAGER
ANALYST EXEC
I..E ALL THE CLERKS TO MANAGERS AND SAME THING WITH ANALYST
OK
NOW I WILL SHOW U ANOTHER QUERY
select job,decode(job,'CLERK','MANAGER','ANALYST','EXEC',JOB) FROM EMP where job=&job;
Enter value for job: 'CLERK'
old 1: select job,decode(job,'CLERK','MANAGER','ANALYST','EXEC',JOB) FROM EMP where job=&job
new 1: select job,decode(job,'CLERK','MANAGER','ANALYST','EXEC',JOB) FROM EMP where job='CLERK'
JOB DECODE(JO
CLERK MANAGER
CLERK MANAGER
CLERK MANAGER
HERE OBSERVE THAT I PASSED JOB ='CLERKS AS PARAMETER AFTER EXECUTING THE QUERY
IT ASKED LIKE THIS AND SHOWED THE REPLACED VALUES
Enter value for job: 'CLERK'
AS SHOWN ABOVE
FOR THIS TO PASS VALUE
IN TO THE QUERY BY USING AMPERSAND('&')
TO TAKE INPUT AS CLERK
IT CHECKED AND RETURNED THE VALUE
IN THE SIMILAR WAY CAN I USE ":" COLON
TO PASS PARAMETERS
IN TO THE QUERY
CAN I USE IT
REGARDS,
PHANI
Edited by: user10652894 on Nov 26, 2008 3:24 AM

Error while replacing IF statements with DECODE function in procedure

Hi All,
I have created a procedure which has nested IF statements. Now I want to replace the IF statements with DECODE functions to improve performance.
Procedure:
IF (var_int_sev = '0')
THEN
var_sev := '2';
ELSE
SELECT sev
INTO var_int_sev
FROM errorconfig
WHERE errorcode = var_errorcode;
var_sev := var_int_sev;
END IF;
I converted the above IF statement into DECODE function as mentioned below:
var_Sev := DECODE(var_int_sev,0,2,SELECT severity FROM errorconfig WHERE errorcode=var_ErrorCode)
But it throws below error at the select statement used inside DECODE.
Error(58,51): PLS-00103: Encountered the symbol "SELECT" when expecting one of the following: ( - + case mod new not null others <an identifier> <a double-quoted delimited-identifier> <a bind variable> avg count current exists max min prior sql stddev sum variance execute forall merge time timestamp interval date <a string literal with character set specification> <a number> <a single-quoted SQL string> pipe <an alternatively-quoted string literal with character set specification> <an alternativ
Can someone help me in converting the IF to DECODE in the above case. Also how can we use a select statement inside decode.

instead of trying to rewrite all your code and hoping that the performance will be better, it's a better option to investigate and find out which part of your application is slow
read this:
When your query takes too long ...

BLOB in DECODE function

Hi All,
I am new to sql. I want to use decode for BLOB datatype. For this i written a query like below.
Here user_data is of type BLOB.
select DECODE(length(user_data), 0, null, file_data) user_data from user_table;
I have tried the other way by using case like below. Its worked fine
select
CASE
WHEN LENGTH(user_data )=0
THEN NULL
ELSE user_data
END user_data from user_table;
What i made wrong in the first query?
Thanks
Jimmi

>
Its just a wrong typo
>
Thanks.
The problem is that this decode
select DECODE(length(user_data), 0, null, file_data) user_data from user_table; expects you to return a number and you are returning a BLOB.
Were you getting 'ORA-00932: inconsistent datatypes: expected BLOB got NUMBER'?
See the DECODE function in the SQL doc
http://docs.oracle.com/cd/B28359_01/server.111/b28286/functions042.htm

Converting Decode function in OBIEE

Hi Guru's
we are converting some BO reports to OBIEE and need to convert Decode function
Below is the Decode function they are using:
Decode ( sign(nvl(BOOKINGS_DATA_UDTC.qty_shipped,0)), 0,decode(BOOKINGS_DATA_UDTC.DELIVERY_NUMBER, 0,decode(BOOKINGS_DATA_UDTC.Schd_Ship_date_fk,to_date('1/1/1990','mm/dd/yyyy'),'Next Month Backlog',decode(sign(nvl(BOOKINGS_DATA_UDTC.Schd_Ship_date_fk, '01-JAN-90') -PAR.PAR_DATE), 1, 'Next Month Backlog', 'Current month will ship')) ,'Awaiting for collection') ,'MTD Shipped')
i have tried converting into case :
case when sign(IFNULL("BOOKINGS_DATA_UDTC"."Qty Shipped",0)) = 0 then (case when "BOOKINGS_DATA_UDTC"."Delivery Number" = 0 then (case when "BOOKINGS_DATA_UDTC"."SCHD_SHIP_DATE_FK" = '01-JAN-1990' then 'Next Month Backlog' else (case when Timestampdiff(SQL_TSI_DAY, cast('@{ParDate}{28-DEC-2012}' as Date),(case when "BOOKINGS_DATA_UDTC"."SCHD_SHIP_DATE_FK" IS NULL then CAst('01-JAN-1990' as DATE) else "BOOKINGS_DATA_UDTC"."SCHD_SHIP_DATE_FK" end)) > 1 then 'Next Month Backlog' else 'Current Month Will Ship' end) end) else 'Awaiting For Collection' end) else 'MTD Shipped' end
But it is not workign as expected.
Can some one please help me with this.
Thanks,

you can achieve it by using CASE WHEN condition END function in obiee. nested case also supported by obiee.
check the below link Decode Join Condition in OBIEE RPD
Thanks
Jay.
Edited by: Jay on Apr 3, 2012 12:56 PM

DECODE function to validate date value and sort the records

Hi Friends,
I am looking for some query which can give me the required output,
I need to do this using SQL query only and I have tried using the MIN() and MAX() functions it was working fine with limited data, now after inserting the last record(in the below table) which has date for start_range and end_range in(mm/dd/yyyy hh24:mi:ss) format.
Because the data type is VARCHAR2 if I am using the MIN() function it is sorting as a string value, hence the min date is incorrect. I tried using validating the value to date or non date and tried to using MIN() and MAX() functions using the DECODE function, I am getting this error "ORA-01840: input value not long enough for date format".
select table_name,
DECODE(substr(START_RANGE,3,1)||substr(START_RANGE,6,1)||substr(START_RANGE,11,1)||substr(START_RANGE,14,1)||substr(START_RANGE,17,1),'// ::',
to_char(min(to_date(start_range,'mm/dd/yyyy hh24:mi:ss')),'MM/DD/YYYY HH24:MI:SS'),min(start_range)) MIN_RUNS_START_RANGE,
DECODE(substr(END_RANGE,3,1)||substr(A.END_RANGE,6,1)||substr(END_RANGE,11,1)||substr(END_RANGE,14,1)||substr(END_RANGE,17,1),'// ::',
to_char(max(to_date(END_RANGE,'mm/dd/yyyy hh24:mi:ss')),'MM/DD/YYYY HH24:MI:SS'),max(END_RANGE)) MAX_RUNS_END_RANGE
from MY_TABLE
GROUP BY table_name,
(substr(START_RANGE,3,1)||substr(START_RANGE,6,1)||substr(START_RANGE,11,1)||substr(START_RANGE,14,1)||substr(START_RANGE,17,1)),
(substr(END_RANGE,3,1)||substr(END_RANGE,6,1)||substr(END_RANGE,11,1)||substr(END_RANGE,14,1)||substr(END_RANGE,17,1))
Can sombody please advise what is the best way I can query this data with the required output.
The following are the source table and records and the out put records using the sql query.
MY_TABLE
TABLE_NAME(VARCHAR2),START_RANGE(VARCHAR2),END_RANGE(VARCHAR2)
TABLE1,1000,10000
TABLE2,ABCD,EEEE
TABLE3,01/12/2010 00:00:00,12/31/2010 23:59:59
TABLE1,10001,20000
TABLE2,EEEF,GGGG
TABLE3,01/01/2011 00:00:00,01/31/2011 23:59:59
OUTPUT :
TABLE_NAME,MIN(START_RANGE),MAX(END_RANGE)
TABLE1,1000,20000
TABLE2,ABCD,GGGG
TABLE3,01/12/2010 00:00:00,01/31/2011 23:59:59
Thanks
Kalycs

i also think this is a very bad table design ...
but if you are not able to change it, you could split the select (date and non-date data) and combine the result with UNION like this:
with t as
SELECT 'TABLE1' table_name,'1000' start_range,'10000' end_range FROM dual UNION
SELECT 'TABLE2','ABCD','EEEE' FROM dual UNION
SELECT 'TABLE3','01/12/2010 00:00:00','12/31/2010 23:59:59' FROM dual UNION
SELECT 'TABLE1','10001','20000' FROM dual UNION
SELECT 'TABLE2','EEEF','GGGG' FROM dual UNION
SELECT 'TABLE3','01/01/2011 00:00:00','01/31/2011 23:59:59' FROM dual
SELECT table_name,
       TO_CHAR(MIN(TO_DATE(start_range, 'MM/DD/YYYY HH24:MI:SS')), 'MM/DD/YYYY HH24:MI:SS'),
       TO_CHAR(MAX(TO_DATE(end_range, 'MM/DD/YYYY HH24:MI:SS')), 'MM/DD/YYYY HH24:MI:SS')
FROM t
WHERE start_range LIKE '%/%/%:%:%'
GROUP BY table_name
UNION
SELECT table_name,
       MIN(start_range),
       MAX(end_range)
FROM t
WHERE start_range NOT LIKE '%/%/%:%:%'
GROUP BY
    table_name;
TABLE_ MIN_VALUE           MAX_VALUE
TABLE1 1000                20000
TABLE2 ABCD                GGGG
TABLE3 01/12/2010 00:00:00 01/31/2011 23:59:59

Is this a bug in decode function?????

Similar Messages

Maybe you are looking for